Middle East Technical University Research Information System
Not a member yet
104897 research outputs found
Sort by
When Ambiguous Loss Becomes Certain Loss: Relatives of Missing Persons in Cyprus
No full textFor more than half a century, the people of Cyprus have reported missing family members to the authorities. Not knowing the fate of a loved one who disappeared during ethnic clashes led families into a state of ambiguous loss, a condition known to cause significant psychological stress. The main goal was to investigate role of identification and funeral process related experience in explaining variance in psychological distress, while controlling for psychosocial variables. Using a cross-sectional design, with total sample size of 898, the study involved two participant groups: Turkish Cypriots (TC) and Greek Cypriots (GC). Separate hierarchical regressions were conducted to examine predictors of psychological distress in both groups. When gender, resilience, perceived injustice, social support, relationship to the missing person (first-degree vs. second-degree), and coping scores were controlled for, experiences related to identification process negatively predicted psychological distress, whereas experiences related to funeral process positively predicted psychological distress in TC group. In contrast, when psychosocial variables were controlled for, experiences related to identification process positively predicted psychological distress, while experiences related to funeral process negatively predicted psychological distress in GC group. These divergent findings regarding identification and funeral processes suggest a potential influence of sociocultural and political factors. These findings highlighted the prolonged psychological distress experienced by both TC and GC family members of identified missing persons. It is considered essential to provide psychosocial support to the families even after the identification process, taking into account cultural, political, religious, and ethnic considerations and needs of each community
Torsional characteristics of a novel infill pattern in material extrusion type of additive manufacturing: random lines
No full textThe selection of an appropriate infill pattern in Fused Filament Fabrication (FFF) plays a critical role in determining both mechanical performance and production efficiency, as it directly influences interlayer adhesion. While various infill patterns have been extensively explored in the literature, this study introduces a novel configuration, referred to as Random Lines, and evaluates its mechanical behavior under torsional loading. The performance of Random Lines is compared against conventional infill patterns-including Lines, Grid, Gyroid, and Archimedean Chord-under identical manufacturing conditions. Standard infill geometries were generated using the open-source slicer Cura, whereas Random Lines and Archimedean Chord patterns were implemented via custom Python scripts. All specimens were fabricated using polylactic acid (PLA), and their mechanical properties were assessed according to the ASTM E143 standard for shear modulus. Given the complex interlayer bonding structures, the polar moment of inertia was determined numerically through image processing to enable accurate conversion of torque and twist angle data into shear stress and strain. The Random Lines pattern demonstrated superior mechanical performance, yielding a shear modulus of 880 MPa-39.1% higher than that of Archimedean Chord-and a yield strength of 26.3 MPa, marking an 85.7% increase. These enhancements, however, were accompanied by a reduction in toughness. Additionally, both Random Lines and Archimedean Chord exhibited smoother toolpaths with fewer retractions and discontinuities, resulting in shorter manufacturing times
Modeling a multi-period mobile parcel locker location-allocation problem with heterogeneous compartments
No full textThe rapid growth of e-commerce and the demand for delivery by last mile has increased the need for sustainable and cost-effective solutions. Parcel lockers have emerged as a promising innovation, yet existing research has rarely addressed mobile lockers with heterogeneous compartments in a dynamic, multi-period setting. This study introduces the Multi-Period Mobile Parcel Locker Location-Allocation Problem with Heterogeneous Compartments (MP-MPLLAP), formulated as a Zero-One Integer Programming model to minimize operational costs, including relocation expenses and customer walking penalties. To handle its computational complexity, a rolling horizon algorithm (RHA) was developed. Empirical data from a parcel locker operator in T & uuml;rkiye were used to test the model. The results reveal a critical trade-off between relocation and customer costs: restricting locker mobility reduces relocation expenses, but increases walking distances, leading to higher overall costs. Furthermore, incorporating heterogeneous compartments proved essential; assuming homogeneity caused capacity overruns while allowing flexible allocation reduced costs at the expense of computation time. The RHA achieved up to 97.3% reductions in computation time and 25.8% reductions in costs on 15 datasets, demonstrating both efficiency and effectiveness. These findings highlight the importance of jointly optimizing relocation and customer walking costs while accounting for compartment heterogeneity in locker planning. The study contributes to theory and practice by providing a realistic, empirically validated model and an efficient heuristic, offering insights for operators and urban logistics planners seeking sustainable last-mile solutions
Optimized Machine Learning Approaches for Parkinson’s Disease Diagnosis via Speech Feature Analysis
No full textParkinson’s disease (PD) is a neurodegenerative condition that severely impairs motor function and reduces quality of life. Effective disease management depends on early and precise diagnosis. Recent advancements in artificial intelligence have significantly enhanced the accuracy of PD diagnosis, particularly through the analysis of speech data. In this paper, we propose a machine learning algorithm based on voice features to classify PD and Healthy Controls (HC). We employed various machine learning classifiers, including Random Forest (RF), Logistic Regression, KNN, Support Vector Machine (SVM), and XGBoost. After hyperparameter tuning, XGBoost achieved the highest accuracy at 96.67%. KNN, RF, and SVM followed with accuracies of 94.87%, and Logistic Regression (LR) recorded the lowest accuracy at 89.74%. Our results demonstrate that effective hyperparameter tuning played a crucial role in achieving higher accuracy compared to similar studies using the same classifiers
Valuing diversity in early childhood education: comparative analysis of New Zealand, Singapore, and Australia curriculum frameworks
No full textActive vibration control of a smart sandwich plate via optimally located piezoelectric sensors and actuators
No full textExcessive vibrations in lightweight structures, such as sandwich composite plates, can lead to performance degradation, fatigue failure, or discomfort in precision applications. Specifically, suppressing the first three coupled modes (including the 1st out-of-plane bending and 1st torsional modes) is crucial, as these modes often form the basis for aeroelastic instabilities like flutter in composite wing structures. To address this challenge, this study presents an active vibration control strategy for suppressing the first three vibration modes of a smart sandwich plate. The proposed method combines a novel curvature-based algorithm for optimal actuator placement with pole placement control for targeted vibration attenuation. The smart structure integrates piezoelectric patches as both sensors and actuators, strategically embedded based on the vibrational characteristics of the base plate. A finite element model and experimental frequency response functions (FRFs) guide the placement and controller design. Ten test cases involving free and forced vibrations validate the approach. Following experimental system identification, transfer functions are derived and used to tune the active controllers. The results demonstrate that the designed control system effectively attenuates each targeted mode, individually and simultaneously, without compromising performance. The study presents a practical and efficient framework for mitigating vibrations in smart structures, thereby contributing to the broader advancement of intelligent adaptive materials in aerospace, automotive, and precision engineering applications
AI-Assisted Arbitrator Selection in Construction Disputes: An Expert-Calibrated Large Language Model Framework
No full textArbitration efficiency is widely recognized as a factor influencing outcomes in construction disputes. To increase the chance of finding and designating the best-fit arbitrator, a large number of candidate profiles must be investigated, which is an overwhelming, time-consuming process. This study develops and evaluates a large language model (LLM)- enabled framework for arbitrator selection based on dispute details and predefined expert criteria. To reach this goal, 500 standardized, anonymized arbitrator resumes were evaluated using a unified scoring structure. These resumes were scored and classified using two GPT-5 models with different levels of detail in their prompts. The results of these models were then compared with expert evaluations to assess their ability to replicate human decision-making patterns in resume evaluation and classification. According to the results, the second model, with a high level of detail in its prompt structure, achieved an accuracy of 84%, while the first model, with a concise prompt that provides only a brief description of the experts’ expectations, achieved an overall accuracy of 53%. As can be concluded, the accuracy of the LLM-assisted resume analysis framework improves when guided by a detailed, expert-aligned prompt structure. From a research perspective, this study’s results highlight the importance of prompt engineering in an AI-assisted decision-support system for professional evaluation tasks. Since this framework is limited to resumes in English, future research should examine the effectiveness of LLMs in evaluating and classifying resumes in languages other than English. Moreover, future studies might consider replicating this study using other large language models to compare precision and accuracy across different LLMs
Bayesian probabilistic photovoltaic power forecasting and stochastic model-predictive control for an agricultural microgrid
No full textThis paper describes a novel probabilistic forecasting method for photovoltaic power for use in energy management for an agricultural microgrid. The forecasting method utilizes recent historical data and a general weather forecast to fit a spline + Gaussian process (GP) model using no-u-turn sampling (NUTS) to infer model parameters in a Bayesian modeling framework. The method seamlessly transitions from a near-term forecast dominated by recent output to a regime where output is dominated by the meteorological forecast. The forecasts are evaluated using proper scoring rules for multivariate probabilistic forecasts and are compared to a reference multivariate persistence forecast and to an LSTM-based forecasting method. The probabilistic method is integrated into a simulation of stochastic model-predictive control (SMPC) for an off-grid agricultural microgrid incorporating photovoltaic generation, a battery storage system, irrigation pumping, and local electrical loads. A 20 %–35 % reduction in simulated operating cost is achieved using the probabilistic forecast compared to a simple expected-value point forecast
Hierarchical human activity recognition with fusion of audio and multiple inertial sensor modalities
No full textIn everyday life, individuals engage in a multitude of activities, and recent technological advancements have facilitated the development of Artificial Intelligence systems that analyse these activities through data in various contexts. Human activity recognition, an essential Artificial Intelligence application in healthcare, detects deviations from normal activities, such as falls, which may indicate health issues. The widespread adoption of mobile and wearable technology enables the development of personalized activity recognition solutions. Given the critical importance of these Artificial Intelligence applications in healthcare, designing systems with high recognition accuracy is imperative. Moreover, these systems must be lightweight to ensure they operate seamlessly on end-user devices like smartphones without compromising their primary functions. Our research introduces innovative input representations and advanced methods in data fusion and multimodal learning that surpass previous methods in recognition accuracy while reducing computational and memory demands. We propose a streamlined neural network model that creatively integrates inertial and audio sensor data to generate color-coded image representations. This implemented Artificial Intelligence system was tested using a complex, publicly available activity recognition dataset organized hierarchically. Compared to earlier studies, our solution demonstrates significant performance enhancements, achieving a 91% balanced accuracy rate in activity recognition and offering substantial improvements in Central Processing Unit and memory efficiency
Effects of biophilic design on sustainable behaviors: introducing the use of serious game as a measure of sustainable behavior
No full textPurposeThis study explores the impact of biophilic design in built environments on sustainable behaviors through the innovative use of a serious game. By examining how exposure to biophilic elements influences behaviors in real and virtual settings, the research aims to demonstrate the potential of serious games as tools for promoting sustainability.Design/methodology/approachThe study was conducted in three distinct experimental settings: (1) a real environment pre-game, (2) a non-immersive game environment within the same real setting and (3) an immersive game environment post-game. Data were collected from 162 participants who experienced these different conditions. The serious game "Pop a Coffee Corner" was developed based on biophilic design principles and used to assess behavioral changes.FindingsResults indicated that exposure to biophilic design elements in real settings significantly enhanced sustainable behaviors compared to non-biophilic environments. Additionally, playing the serious game in a biophilic environment led to even greater improvements in sustainable behavior than exposure to biophilic design alone. This demonstrates the effectiveness of serious games in fostering sustainable actions.Research limitations/implicationsThe study's findings are based on a specific university setting, which may limit generalizability. Future research could explore long-term impacts and applications in diverse contexts.Practical implicationsThe research provides practical guidelines for incorporating biophilic design in built environments, and developing serious games can be a practical strategy for architects, urban planners and educators to promote sustainable behaviors among individuals. This approach can be applied in educational settings, public spaces and workplaces to foster a deeper connection with nature and encourage environmentally responsible behaviors.Social implicationsBy demonstrating the effectiveness of biophilic design and serious games in promoting sustainable behaviors, this study contributes to broader societal efforts to address environmental challenges. Implementing these strategies can lead to increased environmental awareness and pro-environmental behaviors, ultimately supporting sustainability goals.Originality/valueThis study introduces the serious game approach as a novel method to evaluate and promote sustainable behaviors through biophilic design. It highlights the potential for integrating biophilic elements in both real and virtual environments to encourage environmentally responsible behavior, offering valuable insights to architects, designers and policymakers